LAUSR

BACKGROUND We investigated the differential effectiveness of a stepped and collaborative care model (SCM) vs. treatment as usual (TAU) for primary care patients with various depression severity degrees and explored whether subgroups received distinct evidence-based treatments. METHODS Subgroup analyses of a RCT were calculated applying a multiple linear mixed model with the factors 1. group (SCM; TAU), 2. severity ((mild-moderate (MMD); severe depression (SD)) and their interaction, with PHQ-9 as primary outcome. Utilization of treatments was analyzed descriptively. RESULTS For the 737 participating patients (SCM: n = 569; TAU: n = 168), availability of data substantially varies between subgroups at 12-month follow-up ranging between 37% and 70%. ITT-analysis (Last-observation-carried-forward) revealed a significant interaction for group x severity [p = 0.036] and a significant difference between groups in symptom reduction for MMD (-3.9; [95% CI: -5.1 to -2.6, p < 0.001; d = 0.64] but not for SD (-1.6; [95% CI: -3.4 to 0.2, p = 0.093; d = 0.27]. Sensitivity analyses (multiple imputation, completer analysis, pattern mixture model) didn`t confirm the interaction effect and showed significant effects for both severity groups with slightly higher effect sizes for MMD. Differences between SCM and TAU in the percentage of patients utilizing depression-specific treatments are larger for MMD. LIMITATIONS There was a high proportion of missing values among severely depressed patients, especially in SCM. CONCLUSION SCM is effective for both MMD and SD. Utilization patterns might help explain the higher effects for MMD. Various strategies of replacement of missing values lead to slightly divergent results due to selective drop out between severity groups.